CLIMATE CHANGE REPORT - MANITOBA HYDRO'S Insight into the strategies making Manitoba Hydro an industry leader in responding to climate change
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
MANITOBA HYDRO’S CLIMATE CHANGE REPORT Insight into the strategies making Manitoba Hydro an industry leader in responding to climate change Published March 2020
CONTENT INTRODUCTION 6 Climate Change Science 8 Climate Change Indicators 10 Climate Change Strategies 12 1 UNDERSTAND 14 1.1 Global Climate Change 15 1.2 Climate Change & Manitoba Hydro 17 1.3 Manitoba Hydro’s Climate Change Studies 18 1.3.1 Hydroclimatic Monitoring & Analysis 20 1.3.2 Climate Change Scenarios 26 1.3.3 Hydrological Modelling 48 1.3.4 Uncertainty in Future Projections 55 1.4 Memberships, Working Groups, and Research & Development 56 2 REPORT 62 2.1 Emission Trends 64 2.2 Direct Emission Sources 66 2.3 Mandatory Reporting Requirements 69 2.4 Life Cycle Assessments 70 2.5 Reservoir GHG Research 72 1 2
3 REDUCE 74 3.1 Reductions in Fossil-Fueled Generation 75 3.2 Renewable Generation Development 76 3.3 Demand-Side Management 78 3.4 Global Emission Reductions 79 4 SUPPORT 80 4.1 Canadian Climate Mitigation Policies 81 4.2 Manitoba’s Climate & Green Plan 82 4.3 Supporting Climate Policy in Wholesale Markets 83 4.4 Renewable Electricity Energy Markets 84 4.5 Memberships, Working Groups, and Research & Development 85 5 ADAPT 86 5.1 Resource Planning 88 5.2 Streamflow Forecasting 89 5.3 Transmission & Distribution System Reliability 90 5.4 Greenhouse Gas Pricing Implications 91 REFERENCES 92 1 3
MANITOBA HYDRO'S CLIMATE CHANGE STRATEGY CLIMATE CHANGE AFFECTS US ALL International efforts like the Paris Agreement bring humanity closer to solutions, but realities like increased greenhouse gases in our atmosphere and continued global industrialization mean that some climate change effects are inevitable. Our interaction with climate change is two-way: we can be impacted by physical changes in the climate and our operations can impact greenhouse gas emissions that contribute to climate change. • As earth’s climate changes, our environment is also changing, which can affect our water supply, infrastructure, energy demand, and other things. These effects of climate change may require us to adapt in order to continue meeting Manitobans’ energy expectations. • Renewable energy like hydropower is environmentally friendly and helps mitigate climate change. We continually strive to maintain our leadership in reducing global greenhouse gas emissions. 5 STRATEGIES HELP SHAPE OUR RESPONSE 1 UNDERSTAND Earth’s climate is warming and temperature changes affect many of the planet’s natural processes like precipitation, streamflow, and wind patterns. We collaborate with leading researchers and apply rigorously reviewed scientific knowledge to more thoroughly understand historical climate records and future climate projections. 2 AVERAGE AVERAGE MANITOBA REPORT AMERICAN CANADIAN HYDRO’S Canadian governments now require large companies to UTILITY’S UTILITY’S CO2 PER GWh report their greenhouse gas emissions, but we began CO2 PER GWh CO2 PER GWh 0.4 voluntarily reporting in 1995. Reporting on emissions helps us see how we’re doing and how we can improve. 459 135 TONNES TONNES TONNES
WIND 3 REDUCE Greenhouse gas emissions cause climate change. By relying as HYDRO much as possible on renewable resources we are reducing global emissions. The Keeyask hydroelectric generating station is under construction and we have ceased all generation using coal as fuel. Our hydro energy is often abundant and hydropower systems offer flexibility. Exporting extra electricity to neighbouring provinces and states can help reduce their emissions too. 4 SUPPORT 100 + Everyone needs to help reduce greenhouse gas emissions. Through engagement with governments, industry groups, YEARS non‑government organizations, and other stakeholders, we support the development of policies that reduce emissions and help mitigate climate change. Our support also OF RENEWABLE includes openly sharing the technical knowledge ENERGY EXPERIENCE and experience we’ve gained over decades. 5 INCREASED RESILIENCE ADAPT STREAMFLOW FORECASTING Whether it’s increasing our system’s resilience or enhancing our streamflow forecasting capabilities, applying climate information helps us reduce weather-related risks, manage reliability, and capitalize on opportunities. We’re continuing our work to identify, assess, prioritize, and study climate change’s effect on our world and our business. Areas of interest include energy generation, energy demand, environmental sustainability, and infrastructure. WE’RE IN IT TOGETHER Manitoba Hydro is committed to helping the world reduce emissions and mitigate the effects of climate change. As the climate continues to change, we are adapting our processes to ensure we continue delivering reliable, renewable energy to Manitobans.
INTRODUCTION Earth’s climate is dynamic. Changes result from internal (natural) forcing – such as oceanic oscillations – and external forcing such as volcanic eruptions, solar activity, and atmospheric greenhouse gas (GHG) concentrations. The likely dominant cause of recently observed warming since the mid-20th century is human-caused increases in GHG emissions ultimately leading to increased GHG concentrations in the atmosphere [IPCC, 2014]. International efforts like the Paris Agreement bring humanity closer to solutions, but increased GHGs in our atmosphere and continued global industrialization mean some effects of climate change are inevitable. As an energy utility, our interaction with climate change is two-way: physical changes in the climate system impact us, and our operations can impact GHG emissions that contribute to climate change. 1 6 INTRODUCTION
Climate change has been on our radar since the 1980s. This report provides an update to our climate change activities including research, greenhouse gas emission reporting, greenhouse gas reductions, policy support, and adaptation activities. This is the third climate change report, which builds on previous versions (Manitoba Hydro, 2013a; Manitoba Hydro, 2015a). Additional details on some of our past activities, not included in this report, can be found in those reports. Up to now we have focused on “Warming of the climate system during building a foundation of knowledge the Industrial Era is unequivocal, based and tools that enable us to conduct climate change impact studies. We on robust evidence from a suite of employ sophisticated climate models indicators. Global average temperature and downscaling techniques to has increased, as have atmospheric water develop future climate scenarios and examine resulting impacts on business vapour and ocean heat content. Land ice practices. Future water supply has melted and thinned, contributing to remains a large focus of our work sea level rise, and Arctic sea ice has been and these assessments use advanced hydrological models for watersheds much reduced.” of interest to help simulate potential Canada’s Changing Climate Report [ECCC, 2019a] future hydrological conditions. It’s more than water – our impact studies are expanding to help us understand how climate change might affect all our operations. This understanding will help us adapt to continue meeting Manitobans’ energy expectations in the face of climate change. Our low-emitting renewable power also mitigates global climate change. Meeting global climate targets means relying on renewable energy, like hydropower, and the power we export helps mitigate emissions from other forms of energy. We also assist in shaping policy frameworks, guiding development of efficient technologies, and helping domestic customers make wise choices regarding their energy use. INTRODUCTION 1 7
CLIMATE CHANGE SCIENCE To understand how GHGs influence the Earth’s temperature, consider the Earth’s energy balance. It is driven mainly by radiation from the sun. Approximately 30% of sunlight that reaches Earth’s atmosphere is reflected back into space. The rest is absorbed and is converted from light energy into heat. Keeping the energy roughly in balance, Earth radiates heat back to space as longwave (infrared) radiation. GHGs (e.g. water vapor; carbon dioxide CO2; methane CH4; and nitrous oxide N2O) absorb the reflected infrared radiation, acting as a partial blanket. By trapping heat, these gases act like the glass in a greenhouse, warming Earth’s surface. This results in the common name for the GHG-caused climate change: “the greenhouse effect”. This process is critical in maintaining a habitable planet. In the absence of any GHGs, the planet would be too cold to support many life forms. But an excess of GHGs in the atmosphere presents problems. FIGURE 1 Schematic of the greenhouse effect CO2 and other gases in the atmosphere trap heat, keeping the earth warm. Some solar radiation is reflected back into space. 1 8 INTRODUCTION
Burning fossil fuels like coal, natural gas, and oil releases “Human activities are additional heat-trapping gases. Since the industrial estimated to have caused revolution, humans have burned more fossil fuels each successive decade. This intensifies the greenhouse effect, approximately 1.0°C of changing the Earth’s climate. A key atmospheric indicator global warming above pre- is the accumulation of CO2 in the atmosphere. The industrial levels, with a likely average concentration measured at Mauna Loa in 2019 was 411.44 ppm [Trans and Keeling, 2019]. Although range of 0.8°C to 1.2°C. GHG composition within the atmosphere has changed Global warming is likely to over the course of the Earth’s history, the magnitude and reach 1.5°C between 2030 rate of the recent changes appear to be unprecedented. and 2052 if it continues to International efforts to reduce GHG emissions will help limit the impacts of climate change but warming due increase at the current rate. to historic emissions will persist for centuries and will (high confidence)”. continue to cause long-term changes in the climate IPCC Special Report [IPCC, 2018] system [IPCC, 2018]. FIGURE Monthly mean carbon dioxide at Mauna Loa Observatory, Hawaii 2 [Trans and Keeling 2019] Atmospheric CO2 data on Mauna Loa constitutes the longest record of direct measurements of CO2 in the atmosphere. INTRODUCTION 1 9
CLIMATE CHANGE INDICATORS Physical and biological indicators can help us measure changes to the Earth’s environment. Surface temperatures measured on land and at sea for more than a century show a long-term warming trend in globally averaged temperature. The spatial extent of Arctic sea ice is another useful indicator as ice grows and shrinks over the course of the year. Throughout summer, increased solar radiation and higher temperature typically result in sea ice shrinking to its minimum extent each September. Sea ice responds to warmer temperature by retreating further. Minimum sea ice (observed in September) has declined by an average of 13% per decade compared to the 1981–2010 average [Derksen et al., 2018]. Reduced ice and snow cover decreases the amount of sunlight reflected from Earth’s surface and allows for more absorption of heat, which contributes to additional warming. This is an example of “positive feedback”, which reinforces the warming cycle. Studies conducted nationally and regionally have also presented other indicators of a changing climate [ECCC, 2019a; Henderson and Sauchyn, 2008; Lemmen and Warren, 2004; Meehl et al., 2007; van Oldenborgh et al., 2013; Sauchyn and Kulshreshtha, 2008; Warren and Lemmen, 2014]. Some regional studies show that changes that may be attributed to climate change are already being observed in specific components of the environment (Table 1). For example, shifts in seasonality are evident in the observed shortening of the winter season [Vincent et al., 2018], increases in vegetation growth [Ballatyne and Nol, 2015], and decreases in the establishment of perennial lake ice in Northern regions [Paquette et al., 2015]. Seasonal shifts are prominent during the melting season and indicated by earlier ice break up and earlier spring peak streamflow [Bonsal et al., 2019; Derksen et al., 2018; Du et al., 2017]. Due to challenges in attribution studies, longer observational time frames are typically required before more confident statements can be drawn linking detected changes in some indicators to climate change. Climate change studies can often be hindered by short time frames of available observations. Indigenous peoples of regions with observable climate change impacts can provide insights into these challenges through the application of Traditional Knowledge Systems. Traditional knowledge is a separate way of knowing and is the “cumulative body of knowledge, practice, and belief, evolving by adaptive processes and handed down through generations by cultural transmission, about the relationship of living beings (including humans) with one another and with their environment” [Berkes, et al., 2000]. Traditional Knowledge Systems can provide distinct ways of understanding climate change, identifying indicators as well as preparing and applying innovative adaptation techniques. 1 10 INTRODUCTION
TABLE 1 Sample of climate change indicators and references Climate Change Observation Climate Seasonality Indicators Climate Seasonality Shorter winter season Vincent et al., 2018 Permafrost Increasing depth of active Derksen et al., 2018 Thaw Rates layer above permafrost Aquatic Animals Increases in algal production Paterson et al., 2017 & Habitat Decrease in nesting density Ballatyne and Nol, 2015 of whimbrels (shore bird) due Terrestrial Animals to increases in shrub cover & Habitat Increase in Canada goose population Earlier break up of lake ice Derksen, et al., 2018; Du et al., 2017 Unusual loss of perennial Paquette et al., 2015 lake ice cover Lake & Sea Ice Cover Decrease in sea ice extent Kirchmeier-Young, et al., 2017 Decrease in extent of ice and Mudryk, et al., 2018 snow cover Earlier spring peak Bonsal et al., 2019 streamflow Streamflow Higher winter and spring flows INTRODUCTION 1 11
CLIMATE CHANGE STRATEGIES It is clear our activities as humans are resulting in climate change. Reducing GHG emissions and avoiding risks associated with climate change requires a variety of actions to address local and global challenges. We strive to understand and manage risks, liabilities, and opportunities related to climate change. The following five climate change strategies have been established to shape our response to climate change: 1. Understand 2. Report 3. Reduce 4. Support 5. Adapt UNDERSTAND REPORT We strive to understand the Accurately reporting our GHG implications of climate change. This emissions is essential for us to includes maintaining a comprehensive understand our liabilities and to help understanding of the science of us discover opportunities for further anthropogenic climate change, and mitigation. Reporting also allows the the resulting local, regional, and global public and governments to see how hydrological impacts. A comprehensive we are doing, and follow our progress. understanding is vital to ensure Estimating the emissions of our major that we can plan for and adapt to a projects provides transparency and changing physical environment. clarifies their overall impact. 1 12 INTRODUCTION
FIGURE Global implications of Manitoba Hydro operations 3 [Manitoba Hydro, 2019] Net cumulative greenhouse emission reductions. 240 220 Global GHG Emission Reductions 200 (Megatonnes of CO2e) 180 160 140 120 100 80 60 40 20 2004 2008 2016 2000 2002 2012 2018 1998 2006 1994 1996 1992 2014 2010 WIND INCREASED RESILIENCE HYDRO 100 + YEARS OF RENEWABLE STREAMFLOW ENERGY EXPERIENCE FORECASTING REDUCE SUPPORT ADAPT Our operations have always had low We all need effective strategies Responding to a changing physical GHG emissions intensity relative to to mitigate the effects of climate environment means we need other electrical utilities but our entire change and achieve necessary GHG robust plans for potential climate inventory is still emitting at historically reductions. For more than 25 years, scenarios, and we need to position low levels. This is because we have we have lent our technical and market our operations to adapt to changing continued to pursue hydropower, expertise to support the development, parameters like flow conditions and wind generation, and demand-side evaluation, and implementation of electrical loads. We must also adapt to management and removed coal standards, regulations, legislation, the human response to climate change power from our portfolio. Our overall voluntary programs, and markets that that may include changes in societal operations continue to significantly aim to reduce GHG emissions. We preferences for energy sources and help reduce global GHG emissions, far continue to support local, regional, policies and their implications on the outweighing the impact of any direct national, and international climate market price for electricity. emissions we cause. and energy policy dialogues, striving to encourage policies that are both environmentally effective and economically efficient. INTRODUCTION 1 13
1 UNDERSTAND The physical environment is critical to our core business. We continue to invest resources to ensure we understand the changing climate and the potential range of climate change impacts. This positions us to adapt accordingly. 1 14 UNDERSTAND
1.1 GLOBAL CLIMATE CHANGE Understanding the difference between weather The United Nations Environment and climate is critical to studying climate change. Programme and the World • Weather refers to the day-to-day Meteorological Organization variable state of the atmosphere, and is characterized by temperature, precipitation, established the Intergovernmental wind, clouds, and various other weather Panel on Climate Change (IPCC) elements [IPCC, 2013]. Weather results in 1988. The IPCC was created from rapidly developing and decaying weather systems and is challenging to to provide policymakers with predict on a daily basis. regular scientific assessments on • Climate refers to the weather statistics in climate change, its implications terms of its means, variability, extremes, and potential future risks, as well etc. over a certain time span and area as to put forward adaptation [IPCC, 2013]. Climate varies from place to place depending on many factors including: and mitigation options. Along latitude, vegetation cover, distance to large with other reports, the IPCC bodies of water, topography, and other brings together many of the significant geographic features. world’s leading scientists The IPCC refers to climate change when there to prepare comprehensive is a statistically significant variation to the mean state of the climate (or of its variability) that Assessment Reports about the usually persists for decades or longer and which state of scientific, technical and includes shifts in the frequency and magnitude socio-economic knowledge of sporadic significant weather events as well as the slow continuous rise in global mean surface on climate change; the IPCC is temperature [IPCC, 2013]. currently working on their sixth Meeting the Paris Agreement target of holding Assessment Report. Global Mean Temperature (GMT) “well below 2°C” requires an understanding of how GMT has changed historically and is projected to change in the future. The pre-industrial period—until approximately 1750—corresponds to the time before large-scale industrial activity involving fossil fuel combustion and is used as a reference for the 2°C target. However, due to limited observations during this time, 1850–1900 is used by the IPCC to approximate pre-industrial GMT [IPCC, 2018]. The IPCC also commonly reports more recent baseline periods. Kirtman et al., (2013) reported a GMT increase of 0.61°C from the pre-industrial period to 1986–2005. And using the average of multiple datasets, Allen et al., (2018) estimates GMT increases of 0.63°C, 0.64°C and 0.87°C for the 1986– 2005, 1981–2010, and 2006–2015 periods respectively. UNDERSTAND 1 15
“This Agreement... Temperature is not expected to increase uniformly in space or time and there is an interest in understanding impacts at the regional and local scales, but aims to strengthen evaluating historic changes can be challenging due to limited observations. For the global response example, the Hadley Centre – Climatic Research Unit Version 4 dataset [Morice to the threat of et al., 2012] used in Kirtman et al. (2013) and Allen et al. (2018) is gridded at 5° latitude by 5° longitude and has 11 grids that intersect Manitoba. Of these climate change...by 11 grids, the earliest observation is 1872 in the grid containing Winnipeg. holding the increase Considering the 1981–2010 period and using 1872–1900 as a proxy for pre- in the global average industrial times, Winnipeg has warmed approximately 2.3°C. Adjusted and Homogenized Canadian Climate Data (AHCCD) [Vincent et al., 2015] from temperature to Environment and Climate Change Canada suggests 2.6°C of warming over the well below 2°C same periods. Cowtan and Way (2014) applied kriging interpolation to fill gaps above pre‑industrial in the Hadley Centre – Climatic Research Unit Version 4 dataset and show an increase of 1.7°C at Winnipeg from 1850–1900 to 1981–2010, which levels and pursuing highlights some of the uncertainty in quantifying historic changes. efforts to limit the Global Climate Models (GCMs) can also be used to compare projected local temperature increase changes relative to increases in GMT. Following a time sampling approach to 1.5°C above [James et al., 2017] the median of 40 GCM simulations projects 2°C of GMT pre-industrial levels, warming (0.61°C observed plus 1.39°C projected from 1981–2010) to occur in the 2030–2059 period. This future period corresponds to 2.4°C of additional recognizing that this warming at Winnipeg, relative to 1981–2010. Similarly, the average scaling would significantly relationship between GMT increase and local change developed from 40 GCM reduce the risks and simulations project Winnipeg to warm at approximately 1.7 times the rate of GMT which is consistent with supplementary information in Seneviratne et al. impacts of climate (2016). Li et al. (2018) found that Canada as a whole is projected to warm at change” about twice the rate of GMT and shows how several extreme climate indices are The Paris Agreement projected to change. Overall, it is evident that changes in climate can occur more [United Nations, 2015] rapidly at the local scale relative to global average change. FIGURE Examples of weather, natural climate variability, and 4 climate change time scales CLIMATE CHANGE NATURAL VARIABILITY WEATHER HOURS DAYS MONTHS YEARS DECADES CENTURIES Rain Blizzard Wet/Dry El Niño/ Pacific Global Warming Season Southern Decadal Oscillation Oscillation 1 16 UNDERSTAND
1.2 CLIMATE CHANGE & MANITOBA HYDRO Climate change scientists have projected changes in future FIGURE Impacts of climate change on temperature and precipitation patterns, frequency and intensity 5 Manitoba Hydro of severe weather events, and ENERGY SUPPLY sea level rise as a result of rising • Resource availability (water, wind) • Generation planning and operations concentrations of anthropogenic • Financial planning GHGs in the atmosphere [IPCC, • Export markets 2014]. For energy utilities like us, these changes have the potential ENERGY DEMAND (Electricity and Natural Gas) • Decreased winter heating to influence a wide variety of • Increased summer cooling corporate functions (Figure 5). • Policy and technology changes We plan, construct, and operate physical assets based INFRASTRUCTURE DESIGN & MANAGEMENT • Spillways, powerhouses, dykes, transmission on historical climatic and and distribution towers, electrical stations, etc. hydrologic conditions, and • Dam safety and asset management changes in climate may alter • Changing codes and standards their performance. Transmission ENVIRONMENTAL ASSESSMENTS and distribution systems may • Physical environment studies be exposed to a number of • Life cycle assessment and greenhouse vulnerabilities of climate change gas reporting • Stakeholder engagement such as extreme weather events. We are striving to assess HUMAN RESOURCES & CUSTOMER SERVICE the risks associated with climate • Safety and emergency preparedness change and determine how best • Working conditions for field staff • Communication availability to adapt to future conditions. We consider several study domains when conducting climate change studies. Hydrological studies consider all of the basins in the Nelson-Churchill Watershed which supply approximately 97% of our energy in the form of water. This watershed is 1.4 million km2 which covers a sizable portion of central North America and includes a range of different ecozones and geographic areas. The average water volume and energy supplied from each of the major sub-basins is illustrated in Figure 6 as a percentage of the entire Nelson-Churchill Watershed. Other climate change studies, such as those concerning the impact of atmospheric variables on infrastructure, may consider a smaller domain such as the province of Manitoba. UNDERSTAND 1 17
1.3 MANITOBA HYDRO’S CLIMATE CHANGE STUDIES We have initiated a series of comprehensive studies to increase our knowledge of the implications of future climate change. The main objectives of these studies is to incorporate results into long-term planning, operations, and risk management and to adapt infrastructure and business practices to continue serving our core functions. FIGURE 6 Nelson-Churchill Watershed characteristics • Panel A illustrates the major sub-basin names and general flow direction. • Panel B illustrates Ecozones [Commission for Environmental Cooperation, 1997]. 1 A Major sub-basin names and general flow direction 1 B Ecozones Ecozone BOREAL PLAIN HUDSON PLAIN MIXED WOOD PLAINS MIXED WOOD SHIELD SOFTWOOD SHIELD TAIGA SHIELD TEMPERATE PRAIRIES WEST-CENTRAL SEMIARID PRAIRIES WESTERN CORDILLERA 1 18 UNDERSTAND
The approach to these studies is to couple impact models with outputs from reputable climate change modelling centres. We have been working with leading experts (such as those involved in the Ouranos consortium; Section 1.4) in climatology, hydrology, and atmospheric sciences. As new models and tools become available, the ability to project changes in climatic variables at the regional level will evolve. Nelson-Churchill Watershed characteristics (continued)* • Panel C illustrates contribution of total water supply. • Panel D illustrates contribution of total energy supply. Percentages are based on 1981–2010 average inflows available for outflow. For the Churchill River, only a portion of the inflow available for outflow is diverted into the Nelson River. *Totals may not add up to 100% due to rounding 1 C Contribution of total water supply (%) 1 D Contribution of total energy supply (%) 30% 25% 10% 7% 17% 19% 10% 8% 6% 26% 5% 35% UNDERSTAND 1 19
1.3.1 HYDROCLIMATIC MONITORING & ANALYSIS Monitoring & analysis is an important step for characterizing the historical hydrology and climate (hydroclimate) conditions in the Nelson-Churchill Watershed. This information provides the foundation for understanding future hydroclimatic variability and change. MONITORING We monitor changes in the regional climate and hydrology using meteorological and hydrometric information. This information includes measurements of temperature, precipitation, wind speed, and streamflow provided by our Hydrometrics Program Environment, and Climate Change Canada (e.g., Meteorological Service of Canada and Water Survey of Canada), and other gridded and modelled datasets. Under Manitoba Hydro/Manitoba’s Coordinated Aquatic Monitoring Program, we also monitor additional environmental parameters including water quality (more than 50 parameters are analyzed including temperature, dissolved oxygen, pH, etc.), phytoplankton (algae), fish community, benthic invertebrates, and sediment quality. NORMALS (1981–2010) In general, the annual average temperature in the Nelson-Churchill Watershed ranges from −6.5˚C in the northeast to +6.1˚C in the southwest. Total precipitation ranges from 323 mm in the west to 777 mm in the east with some Rocky Mountain regions exceeding 1,000 mm annually. The Nelson-Churchill Watershed shows strong seasonal patterns with colder temperatures and less precipitation in the Climate normals winter, and warmer temperatures and greater precipitation in the represent the average summer. Spring and fall are shoulder seasons with temperature and precipitation normals falling between those observed during winter climatic conditions over and summer. a certain time period at The figures on the following page illustrate annual temperature a certain location. and total precipitation normals across Manitoba and the Nelson‑Churchill Watershed. These annual normals are also supplemented with seasonal normals. Data used to generate these figures is interpolated from observed stations to a 10 km × 10 km grid. A Canadian dataset archived by Natural Resources Canada [Hopkinson et al. 2011, Hutchinson et al. 2009] was merged with a U.S. dataset [Livneh at al., 2013] by Ouranos. Similar to meteorological conditions, hydrological conditions (e.g., annual water supply) within the Nelson-Churchill Watershed is also spatially diverse. 1 20 UNDERSTAND
FIGURE Annual climate normals for average temperature (left) and total precipitation (right) 7 (1981–2010) FIGURE 8 Seasonal average temperature normals (1981–2010) Seasons are winter (DJF), spring (MAM), summer (JJA), and fall (SON). Winter Spring Summer Fall FIGURE 9 Seasonal precipitation normals (1981–2010) Seasons are winter (DJF), spring (MAM), summer (JJA), and fall (SON). Winter Spring Summer Fall UNDERSTAND 1 21
Table 2 summarizes the average streamflow conditions for contributing sub-basins of the Nelson-Churchill Watershed for the 1981–2010 period. Average streamflow near the outlets of each basin varies from 48 cubic metres per second (m3/s) from the Assiniboine River Basin to 947 m3/s from the Winnipeg River Basin. TABLE 2 Average annual streamflow near basin outlets (1981–2010) Annual Streamflow (m3/s) Outlet Basin Station Name Gauge ID Min Mean Max Saskatchewan Saskatchewan 05KJ001 308 551 960 River River at The Pas Assiniboine Assiniboine River 05MJ001 14 48 103 Rivera at Headingley Red River at 05OC001 28 184 405 Red River Emerson Winnipeg River at 05PF063 458 947 1415 Winnipeg Riverb Pine Falls East and West 05UB008 1139 2181 3566 Lake Winnipegc Channels 05UB009 Churchill River 06EB004 574 844 1321 Churchill River above Leaf Rapids Nelson River at 05UF006 2157 3278 5114 Nelson Riverc,d Kettle Rapids a Record reflects losses due to the Portage Diversion b Includes flow from the Lake St. Joseph Diversion c Record represents the combined flow of all upstream basins d Includes Churchill River Diversion 1 22 UNDERSTAND
TRENDS A gridded version of Environment and Climate Change Canada's AHCCD is used to evaluate temperature and precipitation trends. All grids within Manitoban and Canadian portions of the Nelson-Churchill Watershed show statistically significant increasing mean temperature trends over the 1948–2014 period. Statistically significant changes to precipitation were also found in some regions, however, the results are less spatially consistent (Figure 10). Most grids show increasing precipitation trends but decreasing trends can also be found. Despite the variability in precipitation trend direction and magnitude, there seems to be evidence that precipitation has increased in eastern portion of the Nelson-Churchill Watershed. Vincent et al. (2015) presents seasonal trends, examines different time periods, and looks at additional variables such as the snowfall ratio, snow cover, snow depth, and streamflow. FIGURE Historic trends for mean annual temperature (left) and mean annual precipitation 10 (right) (1948–2014) Hatching indicates statistically significant trends at the 5% level. This dataset does not contain information for the United States. Regional and global trends in extreme events are described in the IPCC’s Special Report on Managing the Risks of Extreme Events and Disasters to Advance Climate Change [SREX, 2012] which generally reports greater confidence in temperature- related extremes. Trends in extreme events vary spatially throughout Manitoba. Figure 11 illustrates trends in extreme low temperatures using point values from Environment and Climate Change Canada's AHCCD, for stations with suitable data from 1948 to 2014. Results show increases throughout a large portion of the Nelson-Churchill Watershed with only a few stations reporting insignificant trends. Results are similar for other temperature-related variables including decreases in frost and ice days, increased growing season length, and reduced cold spell durations. UNDERSTAND 1 23
FIGURE Historic trends in extreme minimum temperature using Adjusted and 11 Homogenized Canadian Climate Data (1948–2014) Triangle orientation indicates the direction of the trend. Larger filled triangles indicate statistically significant trends, while smaller open triangles indicate stations where trends are not statistically significant. Colours also indicate trend direction where red shows an increase in temperature and blue shows a decrease. This dataset does not contain information for the United States. Trends in extreme precipitation events are less consistent in space, but show some instances of statistically significant increases in the number of cumulative wet days, and reduced number of cumulative dry days. Some other precipitation- related extreme indices show varied results with increases in some regions and decreases in other areas; and there is low confidence in wind speed-related trends. More complex trends such as multi-year hydrological droughts, with fewer historic events, are more difficult to draw conclusions about. Vincent et al. (2018) provides a more comprehensive view of changes to extreme indices derived from daily temperature and precipitation data. Streamflow trends are useful in representing the area aggregated climate signal within a watershed. However, trend analysis can be challenging due to large natural variability and regulatory effects. Streamflow trends in the Nelson-Churchill Watershed exhibit spatial variability and are sensitive to the time period examined. Using unadjusted Water Survey of Canada data, statistically significant increasing trends in mean annual streamflow were detected for a number of streamflow gauges in the Nelson-Churchill Watershed (Figure 12). Special interpretation is required at some sites due to anthropogenic influences such as diversions in the Winnipeg River Basin and on the Burntwood and Nelson rivers. 1 24 UNDERSTAND
FIGURE Historic trends for mean annual streamflow using unadjusted water survey of 12 Canada data (1975–2014) Triangle orientation indicates the direction of the trend. Large filled triangles indicate statistically significant trends, while smaller open triangles indicate stations where trends are not statistically significant. Colours also indicate trend direction where blue shows an increase in flow and red shows a decrease. This dataset does not contain information for the United States. It is important to acknowledge that trend analysis results can be sensitive to the record length, missing data, and the use of different record periods, all of which can contribute to variability. Trend analysis results are intended to develop an understanding on the direction and significance of historic climate change and are not to be used to project the precise change into the future. ADDITIONAL DATA SOURCES We also use a number of additional data sources to understand historic climate. These data sources include paleoclimate data, oceanic oscillations, reanalysis products from numerical models, and remote‑sensed data from satellites. Paleoclimate data is recognized as a potential source for extending observed records back in time. Sources of paleoclimate data include tree rings and lake sediments which can be correlated to hydroclimatic variables and used as proxy records. We are interested in exploring the use of these datasets in our hydroclimatic studies, but direct applications are currently limited due to availability of long term spatially consistent proxy records and uncertainties in reconstructed hydroclimatic time series, especially in watersheds as large and diverse as the Nelson-Churchill Watershed. We follow the advances of paleoclimatic reconstruction techniques and explore potential applications of paleoclimatic records to better inform our decision making. We also seek to understand connectivity between observed phenomena (such as oceanic oscillations and sun spots) and hydroclimate in the Nelson-Churchill Watershed. While some variability may be explained through study of these relationships, there are challenges in operationalizing the information throughout the entire hydraulic system due to spatial variability and the absence of a single signal that accurately predicts water supply in all hydrological conditions (i.e. wet, dry, and average flow years). UNDERSTAND 1 25
1.3.2 CLIMATE CHANGE SCENARIOS In addition to leading research, compiling information, and providing climate change study guidance, the IPCC also brings together international modelling agencies that have developed GCMs to conduct assessments. The IPCC’s Fifth Assessment Report was released in 2013 and is the most recent report available. Work is Representative Concentration currently underway for the Sixth Assessment Report and it is scheduled for publication in 2021 and 2022. The Fifth Pathways (RCPs) represent Assessment Report was based on results using a suite of GCMs socio-economic and emission from the Coupled Model Intercomparison Project Phase 5 scenarios used as inputs for (CMIP5). Many of the CMIP5 GCMs offer improvements over the previous generation (CMIP3), including finer spatial climate models to explore resolutions and the inclusion of carbon cycling. Simulations plausible future conditions. and analyses are currently underway for CMIP6, which include RCPs consider multiple a number of special experiments such as an abrupt quadrupling of CO2 simulation [Eyring et al., 2016]. factors including population, technology, energy use, and GCMs are numerical models used to translate future atmospheric forcing (e.g. GHG concentrations) scenarios into emissions of greenhouse gases. physically consistent effects on the climate. GCMs compute Four RCPs describe a range energy and mass balances based on physical equations and are of future worlds and their the most advanced tools for projecting future climate. GCM is used herein as a generic term referring to Atmosphere- associated warming potential as Ocean General Circulation Models and Earth System Models. a function of radiative forcing GCMs couple multiple sub-models which simulate various measured in watts per square processes including the atmosphere, ocean, land surface, sea ice, and biosphere. Common variables of interest such as air meter (W/m2). RCP8.5 is a temperature, precipitation, pressure, and wind are products high end scenario (high energy of the atmospheric sub-model. Hydrology is represented intensity, high population coarsely within land surface schemes which output variables such as runoff and soil moisture. growth, limited technology development, and limited GCMs are forced by Representative Concentration Pathways (RCP) [van Vuuren et al., 2011] which are used to prescribe climate policy) and is associated the levels of various forcing agents (e.g. GHGs and aerosols) with an increase of 8.5 W/m2 in the atmosphere. RCPs include a number of assumptions of additional warming on the about societal evolution and represent different demographic, social, economic, regulatory, technological, and environmental earth's surface. developments. Four RCPs are currently considered by the CMIP5 GCMs and they represent a range of futures from the optimistic (RCP2.6) to a business as usual case (RCP8.5). Global CO2 emissions are presently tracking closest to RCP8.5 but given the large time horizon, it is not possible to accurately predict which RCP will be the closest to reality in the year 2100. 1 26 UNDERSTAND
FIGURE 13 Trajectories of CO2 concentration and modelled global surface warming from various Representative Concentration Pathways CO2 GCMs tend to agree on the future warming of the earth however their projection of precipitation and other climatic parameters at the regional or local scale is less consistent and has a greater degree of uncertainty. GCMs use relatively coarse resolutions, ranging from approximately 40 km to 400 km horizontally, and include 18 to 95 vertical levels which can make it challenging to interpret projected changes at finer scales. Therefore, agencies have developed Regional Climate Models (RCMs) which simulate the climate for a limited area such as North America at a finer resolution than the GCMs. Just like the GCMs, these models are physically based but their resolution is typically 50 km or less allowing them to be able to account for important local forcing factors such as better topography representation, especially in mountain regions and other geographic features which GCMs are unable to resolve. Figure 23 and 24 illustrate the resolution of the Canadian Centre for Climate Modelling and Analysis Canadian Earth System Model version 2 (CanESM2) compared to two RCMs. Manitoba Hydro (2015a) employed a GCM simulation ensemble of opportunity (147 simulations from 18 GCMs available at the time) to develop future climate projections. Reflected in this report, the GCM simulation selection process was recently improved to provide a more democratic ensemble [Sanderson et al., 2015]. Future climate projections are based on an ensemble of 40 simulations from 18 GCMs, shown in Table 3. The selection captures GCM simulations with both RCP4.5 and RCP8.5 forcing scenarios that contain monthly output for variables of interest (minimum, maximum, and mean temperature, precipitation, evapotranspiration, runoff, and wind speed) spanning 1981– 2099. The selection process reduces over-representation of GCMs with multiple member UNDERSTAND 1 27
runs and modelling agencies with multiple GCMs. The Kullback-Liebler Divergence [Knutti et al., 2013] was used to guide GCM simulation selection and ensure that a wide range of GCM projection uncertainty was sampled. Figure 14 illustrates the difference between the original 147 simulation ensemble and the new ensemble (sub-set) of 40 simulations. This particular comparison shows that the 40 simulation ensemble forecasts a similar, but slightly warmer and wetter, future compared to the original ensemble on an annual basis. For certain studies, RCMs such as the Canadian Regional Climate Model or the Weather Research Forecast Model are also used and allow analysis at finer spatial resolution. TABLE 3 Global Climate Models [Flato et al., 2013]* Number of Simulations Ensemble Representation Model Country RCP4.5 RCP8.5 by Model Agency BCC-CSM1.1 China 1 1 10% BCC-CSM1.1(m) China 1 1 BNU-ESM China 1 1 5% CanESM2 Canada 2 2 10% CMCC-CM Italy 1 1 5% CSIRO-Mk3.6.0 Australia 2 2 10% GFDL-ESM2g USA 1 1 10% GFDL-ESM2m USA 1 1 GISS-E2-H USA 1 1 10% GISS-E2-R USA 1 1 INM-CM4 Russia 1 1 5% IPSL-CM5a-MR France 1 1 10% IPSL-CM5b-LR France 1 1 MIROC5 Japan 1 1 10% MIROC-ESM Japan 1 1 MPI-ESM-LR Germany 1 1 10% MPI-ESM-MR Germany 1 1 MRI-CGCM3 Japan 1 1 5% *Manitoba Hydro acknowledges the World Climate Research Programme’s Working Group on Coupled Modelling which is responsible for CMIP and the climate modelling groups who produced and made their model outputs available. 1 28 UNDERSTAND
FIGURE Projected changes in annual precipitation and temperature 14 within the Nelson-Churchill Watershed for the 2050s relative to 1981–2010 Each point represents a Global Climate Model simulation with the selected 40 simulations outlined in black. Red circle marker denotes the ensemble average from all simulations while the orange star marker denotes the ensemble average from the sub-set of 40 simulations. Most climate models (GCMs and RCMs) have a tendency to underestimate or overestimate baseline climate conditions. These differences in climate models are called biases when they occur consistently. Applying adjustments to raw climate simulations before they are used in a regional climate analyses is one way we handle these biases. We apply various methods to develop regional scenarios such as dynamic downscaling with a RCM, bias correction with quantile mapping, and the delta method (Figure 15). Bias correction methods aim to adjust the climate simulation time series such that it better matches historic observations while delta methods add the change computed from climate simulations to the observed record [Huard et al., 2014]. The delta method is one of the most common methods as it provides realistic temporal sequencing associated with the historic record and allows future climate change impacts to be evaluated in the context of historical events. UNDERSTAND 1 29
To assist with developing quality regional climate change (downscaled) projections we have become an affiliated member of the Ouranos consortium (Section 1.4). Through its affiliation we gain access to expert guidance for analytical processes used to resolve key features of regional climate and Ouranos’ Canadian RCM data. FIGURE 15 Downscaling methods Global Climate Model Observations/Reanalysis Dynamical Downscaling (regional climate model) Bias Correction Statistical Downscaling (analogue, transfer (delta, quantile mapping) functions, stochastic) Regional Climate Scenarios For the Nelson-Churchill Watershed the GCM ensemble median (using the 40 simulation ensemble of RCP4.5 and RCP8.5) projected changes (deltas) in minimum, maximum, and mean temperature, precipitation, evaporation, runoff, and wind speed for the 2050s (2040–2069) are presented in Table 4, and Figures 16, 17, 19, 20, and 21. Figures present deltas obtained from raw GCM data that have been interpolated to a common grid of 1° latitude by 1° longitude. Future streamflow projections can be found in Section 1.3.3. 1 30 UNDERSTAND
Agreement among GCM projections can provide a measure of confidence. For example, mean annual temperature in the Nelson-Churchill Watershed will likely increase as all GCM projections are in agreement on this direction of change. Some literature refers to this type of information as a measure of robustness or evidence supporting a signal. The IPCC [Mastrandrea et al., 2010] provides guidance on treating uncertainty and suggests qualifiers to express confidence and likelihood (virtually certain; very likely; likely; about as likely as not; unlikely; very unlikely; and, exceptionally unlikely). Since we rely on an ensemble of GCM simulations, we use agreement among these simulations about the direction of projected change to characterize the climate change signal: • strong increase or strong decrease describes signals where 90% to 100% of GCM projections are in agreement; • moderate increase or moderate decrease describes signals where 76% to 89% of GCM projections are in agreement; • weak increase or weak decrease describes signals where 61% to 75% of GCM projections are in agreement; • no signal describes instances where only 50% to 60% of GCM projections are in agreement. These definitions provide a simple means of better understanding certainty in the direction of change but do not provide information about the magnitude of change. The GCM ensemble median is used as a best guess for the magnitude. Projections are presented and discussed separately, by climate variable, below. Projections are tabulated in Tables 4 to 7 and illustrated in Figures 16, 17, 19, 20, and 21. TABLE Global Climate Model ensemble median annual projections for the 2050s 4 relative to 1981–2010 Temperature Wind Watershed Precipitation Evaporation Runoff Min Mean Max Speed Churchill River 3.1°C Ç 2.7°C Ç 2.5°C Ç 7.1% Ç 8.4% Ç 4.0% Ç -0.7% È Saskatchewan River 2.6°C Ç 2.5°C Ç 2.3°C Ç 6.9% Ç 7.8% Ç 6.2% Ç -0.9% È Assiniboine River 2.9°C Ç 2.7°C Ç 2.5°C Ç 7.6% Ç 8.5% Ç -1.2% È -0.5% È Red River 2.9°C Ç 2.8°C Ç 2.6°C Ç 5.6% Ç 7.3% Ç -1.1% È -0.5% È Winnipeg River 3.1°C Ç 2.8°C Ç 2.4°C Ç 6.9% Ç 8.2% Ç 3.3% Ç -0.4% È Lake Winnipeg 3.1°C Ç 2.8°C Ç 2.6°C Ç 7.1% Ç 9.1% Ç 2.3% Ç -0.3% È Nelson River 3.2°C Ç 2.9°C Ç 2.6°C Ç 6.2% Ç 8.0% Ç 1.4% Ç 0.1% Ç Nelson-Churchill Watershed 2.9°C Ç 2.7°C Ç 2.5°C Ç 6.8% Ç 7.8% Ç 4.6% Ç -0.7% È UNDERSTAND 1 31
TEMPERATURE GCMs show a strong signal that temperature will increase in the future at annual and seasonal scales. Annually, the Nelson-Churchill Watershed is projected to experience mean temperatures that are 2.7°C warmer than the baseline. This corresponds to slightly greater changes in the average minimum temperature relative to changes in the average maximum temperature. Seasonally, the Nelson-Churchill Watershed is projected to experience greater temperature increases in the winter relative to other seasons. This is supported in literature suggesting that reduced snow cover in a warmer climate provides lower reflectance (surface albedo) of incoming solar radiation and therefore more absorption of heat by the land surface. Similarly, northern areas are projected to experience greater temperature increases relative to southern areas. One exception to this projection is during summer months where southern areas may experience slightly greater increases than northern areas. This behaviour is possibly due to the absence of precipitation which reduces capacity for evaporative cooling in the summer. TABLE Global Climate Model ensemble median 5 seasonal temperature projections for the 2050s relative to 1981–2010 Watershed Winter Spring Summer Fall Churchill River 3.8°C Ç 2.4°C Ç 2.2°C Ç 2.8°C Ç Saskatchewan River 3.1°C Ç 2.1°C Ç 2.5°C Ç 2.5°C Ç Assiniboine River 3.5°C Ç 2.3°C Ç 2.6°C Ç 2.8°C Ç Red River 3.5°C Ç 2.3°C Ç 2.6°C Ç 2.7°C Ç Winnipeg River 3.4°C Ç 2.4°C Ç 2.6°C Ç 2.7°C Ç Lake Winnipeg 3.6°C Ç 2.5°C Ç 2.5°C Ç 2.8°C Ç Nelson River 3.9°C Ç 2.6°C Ç 2.3°C Ç 2.9°C Ç Nelson-Churchill Watershed 3.4°C Ç 2.3°C Ç 2.4°C Ç 2.7°C Ç 1 32 UNDERSTAND
FIGURE Global Climate Model ensemble median projected change (left) and agreement (right) for 2050s 16 mean annual temperature relative to 1981–2010 (top panels) Seasonal projected change and agreement shown in lower panels. Seasons are winter (DJF), spring (MAM), summer (JJA), and fall (SON). SEASONAL PROJECTED CHANGE Winter Spring Summer Fall SEASONAL AGREEMENT Winter Spring Summer Fall UNDERSTAND 1 33
PRECIPITATION Annually, the GCM ensemble shows a moderate to strong signal that annual precipitation will increase in the Nelson-Churchill Watershed. The ensemble median projects a spatially averaged 6.8% increase. The GCM ensemble shows moderate to strong signals that precipitation will increase in winter and spring accompanied with weak to strong signals that precipitation will increase in fall. Southern (northern) basins project decreasing (increasing) summer precipitation but the agreement is weak. TABLE Global Climate Model ensemble median 6 seasonal precipitation projections for the 2050s relative to 1981–2010 Watershed Winter Spring Summer Fall Churchill River 11.3% Ç 10.5% Ç 4.7% Ç 8.1%Ç Saskatchewan River 9.2% Ç 14.3% Ç -0.4% È 5.6%Ç Assiniboine River 9.2% Ç 16.8% Ç -1.7% È 8.3%Ç Red River 7.7% Ç 11.7% Ç -0.8% È 5.0%Ç Winnipeg River 11.3% Ç 11.5% Ç -0.1% È 5.7%Ç Lake Winnipeg 11.3% Ç 12.2% Ç 0.0%ÅÆ 7.6%Ç Nelson River 12.8% Ç 10.7% Ç 2.1% Ç 6.5%Ç Nelson-Churchill Watershed 10.9% Ç 14.0% Ç 0.1% Ç 7.8%Ç 1 34 UNDERSTAND
FIGURE Global Climate Model ensemble median projected change (left) and agreement (right) for 2050s 17 mean annual precipitation relative to 1981–2010 (top panels) Seasonal projected change and agreement shown in lower panels. Seasons are winter (DJF), spring (MAM), summer (JJA), and fall (SON). SEASONAL PROJECTED CHANGE Winter Spring Summer Fall SEASONAL AGREEMENT Winter Spring Summer Fall UNDERSTAND 1 35
EVAPORATION The GCM ensemble shows moderate to strong signals that evaporation will increase in the future annually as well as in winter and spring. Increases are also projected for summer and fall, but there are regions of less agreement (weaker signals; Figure 19). This is largely due to evaporative potential being driven by temperature and precipitation. FIGURE January morning evaporation near the Long Spruce 18 generating station Despite frigid air temperatures, evaporation can still occur in winter. When cold, dry air blows over a (relatively) warm water body, it is heated by the water’s surface and humidified through evaporation. This relatively warm air quickly cools as it rises from the water surface, causing the water vapour in the air to condense into a thick fog. 1 36 UNDERSTAND
FIGURE Global Climate Model ensemble median projected change (left) and agreement (right) for 2050s 19 mean annual evaporation relative to 1981–2010 (top panels) Seasonal projected change and agreement shown in lower panels. Seasons are winter (DJF), spring (MAM), summer (JJA), and fall (SON). Some grids are masked due to differences in GCM land surface schemes and representation of large water bodies like Lake Winnipeg. SEASONAL PROJECTED CHANGE Winter Spring Summer Fall SEASONAL AGREEMENT Winter Spring Summer Fall UNDERSTAND 1 37
RUNOFF GCM runoff is used as a basic measure of water availability [Frigon, 2010] to better understand changes in water supply. Due to limitations in GCM representations of hydrological processes (e.g., coarse resolution and lack of routing) GCM runoff is used as a preliminary variable, providing a broad view of how runoff is projected to change over large geographic areas. GCM runoff projections are complemented with more thorough hydrological modelling to examine finer details such as seasonal shifts in timing at finer temporal and spatial resolutions. This is covered in Section 1.3.3. The GCM ensemble median projects a 4.6% increase in mean annual runoff in the Nelson-Churchill Watershed, but there is little agreement among GCM simulations in most areas. Some northern and eastern parts of the watershed show weak to moderate agreement that runoff will increase annually and a small area in the south shows weak agreement that runoff will decrease. Similar signals are seen for summer and fall. In contrast, a strong signal is seen throughout a majority of the basin indicating that winter runoff will increase. This is likely due to warmer temperatures reducing the duration when precipitation is stored as snow and not contributing to runoff. Warmer temperatures may also lead to increased rain-on-snow events and snowmelt which contribute to runoff. GCMs also show weak agreement in a large portion of the watershed that spring runoff will decrease. This behaviour can likely be attributed to reduced snowpack accumulated over the winter or increased temperatures causing more evaporation. It is important to note that for large basins, seasonal runoff changes may not directly correspond to streamflow changes in the same season as there is often a lag due to river routing, lake attenuation, and regulation. 1 38 UNDERSTAND
You can also read